Molecular divergence of maintainer and restorer lines of tropical rice hybrids

Thirty-seven SSR markers distributed in 12 rice chromosomes were selected for the molecular diversity assay of 37 maintainer lines (B) from the International Rice Research Institute (IRRI) and 44 restorer (R) lines from the Philippine Rice Research Institute (PhilRice). The SSR assays followed the standard procedure described by Panaud et a!. (1996). Nei and Li’s statistic (Nei and Li 1979) was used as a measure of genetic similarity. Gene diversities between different lines were calculated using algorithm: Hk = 1-sigma P,2. given Pi is the frequency of the ith allele at kth locus. Total gene diversity was partitioned into its components in a manner: HT= Hb + Hr + HBR, where HT, Hb, HR, and HER are weighted gene diversity in total, within B population, within R population and between B and R populations, respectively. The corresponding relative gene diversities are given by dividing each component with HT.

The mean number of alleles per SSR locus was 4.24 ± 1.71, ranging from 2 to 9 (Table 1). There were 14 loci where more than 5 alleles were detected. Out of 157 detected alleles, 115 (73.2%) alleles were common to both B and R lines, 11 alleles at 9 loci were unique to B lines, while 3 lelleles at 14 loci were unique in R lines. The frequency of group specific alleles ranged from 0.02 to 0.16. Relative gene diversity averaged 0.20 within B lines, 0.28 within R lines, and 0.52 between two groups of lines, suggesting a significant divergence between the parental groups.

Significant allelic frequency differences between B and R groups were present for the two most common alleles at the majority of the SSR loci. The mean genetic distance was 0.39 between pairwise B lines, and 0.49 between R lines, indicating the B lines shared greater similarity within group than that of the R lines. This is understandable because among the tropical indica elite lines frequency of B lines (less than 5%) is much lower than that of R lines (20-30%) for the WA-CMS system.

Cluster analyses were performed based on Nei’s genetic distances between lines within the B and R groups (Fig. 1 and 2). The B lines formed 4 clusters. Cluster A consisted of 33 cultivars (87%). Cluster B contained 3 closely related lines, B38, B40, and B 16. Cluster C contained only a single line, B39, which was known to have wide compatible genes from BPI 76 and Palawan. The line, B2 formed another single line cluster, D. The 44 R lines formed three clusters with cluster A containing 41 most recently developed advance lines. Cluster B consisted of only two lines, R62 and R7 derived from crosses of Ostralia/ Pinilian and BE-3IBPI 121, respectively. The line, R86 formed a single line cluster.

Our results indicated a clear differentiation between the B and R lines in tropical hybrid rice breeding programs. As expected, the genetic diversity within the R lines was more pronounced than within B lines. Future research efforts should be devoted to increase diversity among B lines and between B and R lines in order to achieve a broad genetic basis of rice hybrid.

References
Chen, X., S. Tenmkh, Y. Xu, Y.G. Cho and S.R. McCouch, 1997. Development of a microsatellite framework map providing genome-wide coverage in rice (Oryza sativa L.). Theor. Appi. Genet. 95: 553-567.